Alexithymia

The alexithymia concept derived from clinical observations that psychosomatic patients were unimaginative and showed difficulties with the verbal and symbolic expression of emotion (Ruesch, 1948). Later, empirical studies demonstrated this phenomenon. For example, Sifneos (1973) found that psychosomatic patients tended to have a relative constriction in emotional functioning, a poverty of fantasy, and an inability to find appropriate words to describe emotions. Psychosomatic patients possessed twice as many alexithymic characteristics when compared to controls.

Recently, Sifneos expanded on the concept, “Alexithymia, a term I introduced for better or worse in 1972, involves a marked difficulty to use appropriate language to express and describe feelings and to differentiate them from bodily sensations, a striking paucity of fantasies and a utilitarian way of thinking which Marty et al [1963] have called penseé opératoire” (Sifneos, 2000).

According to Taylor, Doddy & Newman 1981 ‘alexithymia’ refers to a hypothetical personality construct that is characterised by, “(1) a difficulty in identifying and communicating feelings, (2) a difficulty in distinguishing between feelings and bodily sensations (3) impaired symbolization, as evidenced by paucity of fantasies and other imaginative activity, and (4) a preference for focussing on external events rather than inner experiences.” (Taylor & Bagby 1988, p.352).

As a reflection of the increasing interest in this construct, alexithymia was chosen as the main theme of the 11^th European Conference on Psychosomatic Research at Heidelberg, in 1976 (Brautigam & Von Rad, 1977). Indeed, Taylor in 2000 comments that, “While about 120 articles were published on alexithymia by the mid-1980s, a recent search of the PsycInfo database revealed well over 700 journal articles on alexithymia.” (Taylor, 2000, p.134.).

How is it measured?

The Beth Israel Hospital Psychosomatic Questionnaire (BIQ)

This measure of alexithymia is a 17-item forced choice questionnaire that was developed by Sifneos to measure psychosomatic diseases. The items are completed by the interviewer after a structured interview with the patient. Eight of the items here are thought to be directly related to the alexithymia construct, and therefore only these 8 items are used in obtaining an alexithymia score. A cut-off score, arbitrarily chosen, of 6 yields an alexithymic (≥6) and non-alexithymic (<6) categorisation. There have, however, been a number of methodological criticisms of this scale. The forced choice nature of this scale (true, false) may be unsatisfactory as dichotomous scales can be unstable in factor analyses (Taylor & Bagby, 1988). Perhaps the most pertinent criticisms levelled at the BIQ are that it is time-consuming, and subject to experimenter or observer bias (Kleiger & Kinsman, 1980; Lolas, De La Parra, Arohnson & Colin 1980; Taylor, Doddy & Newman 1981).

Schalling-Sifneos Personality Scale (SSPS)

This self-report scale has 20 items and uses a 4-point Likert scale. Lower scores indicate increased levels of alexithymia. Sifneos suggests that a score of 50 or below indicates alexithymia. Unfortunately, this test was not subjected to item analysis during its construction, leaving it psychometrically weak. Other studies have found that its internal consistency is unsatisfactory (Bagby, Taylor & Ryan 1986; Bagby, Taylor & Atkinson 1988; Faryna, Rodenhauser & Torem 1986). Parker, Taylor & Bagby (1991) have strongly criticised a revised version of the SPSS (Sifneos, 1986) on psychometric grounds, and Bagby, Taylor & Atkinson (1988) advise against its further use.

The MMPI (Welsh & Dahlstrom, 1963; Good & Brantner, 1974) is a psychometric instrument that has been used to assess personality traits in a variety of medical and psychiatric syndromes. It contains 566 dichotomous choice (true or false) items that relate to mood, behaviour, self-concept and personal preferences. An alexithymia scale on the MMPI was developed (Kleiger & Kinsman, 1980) in response to a finding that the MMPI did not distinguish between alexithymic and non-alexithymic groups as measured by the BIQ (Kleiger & Jones, 1979). As Parker et al (1991) have pointed out, the alexithymia scale on the MMPI has been criticised for a lack of internal consistency, having a social desirability bias, and not relating well to other constructs. As with the SSPS, Bagby et al (1988) advise against future use of the MMPI-A.

The Toronto Alexithymia Scale (TAS)

The TAS was constructed after a literature review revealed 5 main content areas thought to reflect the construct. Forty-one items were devised based on the content areas. From these 41 items, 15 were deleted for failing to meet 2 statistical criteria. This left 26 items from which a factor analysis suggested a four-factor solution. Factor I consisted of items that refer to the ability to identify and describe feelings and to distinguish between bodily sensations, factor II reflected the ability to communicate feelings to others, factor III represented the ability to daydream, and IV represented the tendency to focus on external events over inner experiences. The TAS uses a 5-point Likert type rating scale from 1 (strongly disagree) to 5 (strongly agree). The TAS has shown adequate internal consistency, good test-retest reliability, and good convergent and discriminant validity (Taylor et al, 1997). However, there are some limitations to the TAS.

The ‘day dreaming’ factor has been shown to correlate negatively with alexithymia and has been partly explained by a social desirability response bias. There have also been high correlations between factors I and II and very low correlations between factors III and IV (Taylor et al, 1997). In response to these weaknesses and in recognising that scale development is an ongoing process, further revisions were made.

The Twenty Item Toronto Alexithymia Scale (TAS-20)

After a revised edition to the TAS, a twenty-item version was devised. The TAS-20 has 3 factors including: difficulty identifying feelings and distinguishing them from bodily sensations (F1), difficulty describing feelings to others (F2), and externally oriented thinking (F3). Preliminary evidence of reliability and factorial validity has been established (Bagby, Parker & Taylor, 1994-II). The TAS and TAS-20 are now the most widely used measures of alexithymia (Taylor, 2000).

Relationship with somatic illness

It is proposed that the limited emotional awareness and cognitive processing of affect seen in alexithymia, leads to individuals focussing on and amplifying the somatic aspect of emotional. This may explain the apparent association between alexithymia and psychiatric disorders with somatic presentations and even somatic illness. Alexithymia has been associated with hypertension (Todarello, Taylor, Parker & Fanalli 1995), inflammatory bowel disease (Porcelli, Zaka, Leoci, Centonze, Taylor & Parker 1995), functional gastrointestinal disorders (Porcelli et al, 1999), somatoform disorders (Cox, Kuch, Parker, Shulman & Evans 1994), panic disorder (Zeitlin & McNally, 1993) and eating disorders (De Groot, Rodin & Olmstead, 1995).

Emotional processing and alexithymia

In summary, alexithymia is a hypothetical personality construct that involves impairments in identifying feelings and describing feelings, a paucity of fantasy life, and a tendency towards externally oriented thinking. How then, does alexithymia relate to emotional processing?

A recent empirical study outlining the psychometric properties of a new emotional processing scale (EPS) showed that alexithymia is related to emotional processing (Baker, Thomas, Owens & Thomas 2003 in preparation).

Correlations between the TAS-20 subscales and the EPS total score are quite high (identifying feelings r=0.69, describing feelings r=0.67, and the total TAS-20 r=0.71). The externally oriented thinking subscale produced a more moderate correlation with total EPS (r=0.30). This suggests that there is some relationship between the two constructs (approximately, a 50% overlap). Intuitively, this seems to make sense as the two constructs pertain to both emotions and to cognitive-emotional processes. Exploring the data of the study further, there were low correlations between the TAS-20 total score and some of the EPS subscales. For example, on the intrusive and persistent thoughts subscale (r=0.22), the can’t control subscale (r=0.36), and the avoidance subscale (r=0.39). It would appear that these subscales of the EPS are measuring dimensions other than alexithymia.

Emotional processing was designed as broader in scope than alexithymia. It was meant to apply particularly to patients with psychological problems, physical illness and psychosomatic conditions, as well as healthy or 'normal' states of mind. Its aim is to capture the various processes at work in emotional processing at psychological, psychoneurological and physiological levels, rather than refer to a personality trait, a type of individual or a diagnostic condition or category. In emotional processing a range of different deficits are conceivable, for instance patients with anxiety disorders may show a different pattern of emotional processing to those with depression. Alexithymia is concerned with one trait or category, rather like a diagnosis. Its roots are psychoanalytic and medical; the emotional processing model is more closely related to the concepts of clinical psychology.

Alexithymia network:

http://alexithymia.med.up.pt/public/docs/rightframe_files/books-body.htm

References

Dorset RDSU
Cornelia House Poole Hospital NHS Trust Longfleet Road Poole Dorset BH15 2JB
Email dawn.stevens@poole.nhs.uk